Algorithm 1: Pseudo code of the seed cluster algorithm
	Input: Pxyz (points), Pfsp (feature space parameter of points)
	Output: PSCxyz (seed cluster points)
1	nob = initial number of bins
2	for i=1 to 4 do
3	Compute histograms of Pfsp with nob bins
4	db (dominant bin) = {binm ∈ histogram: |binm| = max(|binn|), n ∈ nob}
5	Pfsp {P: Pfsp ∈ db}
6	Eliminate feature space parameter containing db from Pfsp
7	end
8	if |Pfsp| >= 5 then
9	PSCxyz = Pfsp
10	else if nob < 4 then return Ø
11	else nob = ceil(n/2); JUMP TO 2